ABSTRACT
OBJECTIVES: Cyclooxygenase 2 (COX-2) and vascular endothelial growth factor (VEGF), often coexpressed in cancer, are associated with poor prognosis. However, results from pancreatic cancer trials of their inhibitors were disappointing. This study delineated the role of COX-2 and nonsteroidal anti-inflammatory drugs in angiogenesis and VEGF regulation. METHODS: AsPC-1 and BxPC-3 pancreatic cancer cells were cocultured with human umbilical vein endothelial cells (HUVECs). NS398 or VEGF-neutralizing antibody was added, and HUVEC viability assayed. Prostaglandin E2 and VEGF were quantified. Tumor cells were treated with NS398 or celecoxib, and VEGF quantified. RESULTS: In cocultures, HUVEC viability in AsPC-1 was 60% that of BxPC-3 controls (P < 0.05). Prostaglandin E2 and VEGF from BxPC-3 were double that of AsPC-1 (P < 0.05). NS398 reduced prostaglandin E2 to undetectable levels (P < 0.05) but had no effect on HUVEC viability. Vascular endothelial growth factor-neutralizing antibody reduced HUVEC viability in BxPC-3 wells to that of AsPC-1 (P < 0.05). NS398 had no effect on VEGF. Celecoxib increased VEGF in a concentration-dependent manner in each cell line up to 4-fold (P < 0.05). CONCLUSIONS: Cyclooxygenase 2 does not regulate VEGF in pancreatic cancer, and celecoxib upregulates VEGF in pancreatic cancer. It is VEGF, and not COX-2, inhibitors that reduce tumor-stimulated endothelial cell viability. Future pancreatic cancer trials should consider lower-dose nonsteroidal anti-inflammatory drugs in combination with VEGF inhibitors.
Subject(s)
Cyclooxygenase 2/physiology , Endothelium, Vascular/metabolism , Neovascularization, Pathologic/metabolism , Pancreatic Neoplasms/blood supply , Vascular Endothelial Growth Factors/physiology , Angiogenesis Inhibitors/pharmacology , Angiogenesis Inhibitors/therapeutic use , Anti-Inflammatory Agents, Non-Steroidal/pharmacology , Anti-Inflammatory Agents, Non-Steroidal/therapeutic use , Celecoxib , Cell Survival/drug effects , Cyclooxygenase 2 Inhibitors/pharmacology , Cyclooxygenase 2 Inhibitors/therapeutic use , Dinoprostone/analysis , Endothelium, Vascular/drug effects , Endothelium, Vascular/pathology , Humans , Neovascularization, Pathologic/drug therapy , Neovascularization, Pathologic/pathology , Pancreatic Neoplasms/metabolism , Pyrazoles/pharmacology , Pyrazoles/therapeutic use , Sulfonamides/pharmacology , Sulfonamides/therapeutic use , Tumor Cells, Cultured , Vascular Endothelial Growth Factors/antagonists & inhibitorsABSTRACT
Dendritic cells (DCs) are critical to the outcome of many viral infections. Questions still remain as to the relevance of artificially generated DCs in models of in vivo immune responses. We compared different DC generation pathways, in terms of phenotypic expression, cytokine production, apoptosis, and T cell proliferation, following viral infection. Direct viral infection of monocytes or monocytes cultured with supernatants from virally infected lung epithelial cells (A549 DCs) induce distinct DC subsets compared with viral infection of artificially generated IL-4 DCs and IFN-DCs. These virally infected DC subsets stimulated different cytokine secretion profiles and displayed contrasting sensitivities to viral-induced apoptosis. It is most interesting that we observed marked differences in the proliferation of purified CD3+ T cells from the virally infected DC subsets. In conclusion, artificially generated DCs skew immune responses to viral infections, and direct viral infection of monocytes and DCs, generated from monocytes cultured with supernatants from infected epithelial cells, appears to be a more relevant pathway of producing DCs, which mimic those generated in vivo.
